[evlatests] OTF Holography Tests

[Bryan Butler]

Take the difference of an OTF'ing antenna with one of the reference 
antennas (10, 23, 28) during the OTF beamcuts.  That will give you 
delta-az, delta-el, as a function of time, which can be transformed to 
(l,m).  Parse what you want out of the Pointing.xml file where the 
modified SDM is (the directory I sent you); I used Rich's converter to 
get XML from binary.  There is a Perl script there (readPtgSDM_2.pl) 
that will parse out a single antenna and print the relevant values to 
STDOUT.  It takes a while (about 15 minutes per antenna) because 
Pointing.xml is quite large.  A couple of examples are in antenna_2.out 
and antenna_10.out (I didn't chop out just the times on the OTF beamcuts 
- it's all in there for the whole run).  I may have to add more 
precision for this particular application.  Bill would have to add 
similar mechanics to BDFIN if you wanted this all done automagically, 
I'd guess.

	-Bryan


Ken Sowinski wrote on 2/11/22 08:55:
> On Thu, 10 Feb 2022, Bryan Butler via evlatests wrote:
>>
>> Almost certainly easiest to get it from the Pointing table in the SDM.
> 
> I have been trying to convince myself that the Pointing table
> contains all the necessary information to implement Barry's second
> suggestion, but have not yet succeeded.
> 
> Ken
> 
> 
>>     -Bryan
>>
>>
>> Barry Clark via evlatests wrote on 2/10/22 16:29:
>>>  The l,m can be easily calculated from the time and the reference 
>>> time in
>>>  the script.  Easiest way is probably to do the calculation and store 
>>> the
>>>  results in the u,v slot.  Problem is knowing what the reference time 
>>> was
>>>  and where the antenna was commanded to be pointed at the reference 
>>> time.
>>>  These can be grepped from the .evla script, but that's pretty 
>>> inelegant.
>>>  Alternative is to calculate true az-el of the source for each time, and
>>>  subtract from the antenna position stream - probably easier in CASA 
>>> than
>>>  AIPS.
>>>
>>>  On 2/10/2022 3:49 PM, Rick Perley via evlatests wrote:
>>>>  A few weeks back, Ken requested some help in testing out a new 'OTF'
>>>>  scanning mode -- to be used for antenna holography.  This 
>>>> initiative was
>>>>  spurred by suggested holography measurements for the ngVLA prototype
>>>>  antenna.
>>>>
>>>>  Last week, a short test was done to see if Ken's initial stab at 
>>>> this was
>>>>  in the right direction.  I generated an SB with two orthogonal cuts 
>>>> using
>>>>  our traditional 'stop and stare' method. The test had 31 positions, 
>>>> each
>>>>  10 seconds, done at X-band on 3C84. 5X oversampling was used (i.e., 
>>>> the
>>>>  grid spacing was lambda/5D radians).
>>>>
>>>>  To this Ken added two paired orthogonal OTF cuts:  The first pair 
>>>> with a
>>>>  scanning speed set to 'match' the stepped test, the second pair to go
>>>>  twice as fast.
>>>>
>>>>  The usual difficulties expected when trying something new were
>>>>  experienced, but with Bill Cotton's help, a data file readable by 
>>>> BDFIn
>>>>  was produced.
>>>>
>>>>   Attached are ten UVPLT files -- no calibration was done.  Single
>>>>  baseline (fixed x moving:  ea28 x ea01) single polarization.  No
>>>>  averaging.
>>>>
>>>>  The first two (#1 and #2) show the standard 'stepped' profiles. 
>>>> First in
>>>>  Elevation, second in Azimuth.  (The overshoot on the steps show 
>>>> that ea28
>>>>  has an old ACU).
>>>>
>>>>  The last eight show Ken's OTF cuts.   The first four are each 95 
>>>> seconds
>>>>  long, the last four are half that (twice as fast).  By comparing the
>>>>  heights of the sidelobes, I can tell that:
>>>>
>>>>  #3 and #7 plots are in azimuth, reversed direction from #1.
>>>>
>>>>  #4 and #8 are in azimuth, same direction as #1.
>>>>
>>>>  #5 and #9 are in elevation, reversed direction from #2
>>>>
>>>>  #6 and #10 are in elevation, same direction as #2.
>>>>
>>>>  The profiles are just gorgeous!  The motion is clearly smooth and
>>>>  continuous.  No sign of problems at beginning or end.  The 
>>>> 'double-speed'
>>>>  profiles look identical to the 'half-speed' ones. Phases look good as
>>>>  well.
>>>>
>>>>  That's the good news.
>>>>
>>>>  The bad news is that the (l,m) coordinates are rubbish.  All zeros for
>>>>  the first seven 'OTFRASTER' scans, and some enormous numbers for that
>>>>  last scan.
>>>>
>>>>  Ken is now thinking about how to get proper (l,m) coordinate 
>>>> information,
>>>>  and in adding various other options.
>>>>
>>>>
>>>>  _______________________________________________
>>>>  evlatests mailing list
>>>>  evlatests at listmgr.nrao.edu
>>>>  https://listmgr.nrao.edu/mailman/listinfo/evlatests
>>>
>>>  _______________________________________________
>>>  evlatests mailing list
>>>  evlatests at listmgr.nrao.edu
>>>  https://listmgr.nrao.edu/mailman/listinfo/evlatests
>>
>>
>> _______________________________________________
>> evlatests mailing list
>> evlatests at listmgr.nrao.edu
>> https://listmgr.nrao.edu/mailman/listinfo/evlatests
>>
>>